Table shows the results of the multivariable linear
Table 2 shows the results of the multivariable linear regression. Briefly, in crude analyses (adjusting only for urinary creatinine) individuals in the 3rd and 4th quartiles of urinary MEHP had statistically significantly longer LTL compared to the lowest referent quartile (Table 2, Model 1) with evidence of a dose–response relationship (p-value for trend <0.001). Further adjustments for demographics (among them age) and socio-behavioral variables (Table 2, Model 2) and for body weight status and c-reactive protein (Table 2, Model 3) showed decreased parameter estimates for MEHP compared to Model 1, likely driven by the role of aging in telomere attrition and the other covariates such as body weight, but the statistical significant association of MEHP with longer LTL remained. There were no statistically significant associations of the phthalate metabolites MEP, MBP and MBzP with LTL (Table 2).
In complementary analyses, using urinary phthalate metabolites as natural log-transformed continuous variable of the association of increased urinary MEHP level with longer LTL was confirmed (Table 3). With each one-unit of natural log-transformed MEHP unit, there was a 1.71% increase in LTL (Table 3, Model 3). Furthermore, analyses using restricted cubic spline confirmed the dose-relationship between MEHP and LTL (Supplement Fig. 1).
Analyses stratified by age group indicated that arn-509 the statistically significant association of LTL with MEHP was found in young adults (20–39years), middle aged (40–59years) and older adults groups. In the young adult group there was statistically higher LTL percentage change in the 2nd, 3rd and 4th MEHP quartiles compared to the lowest MEHP quartile, with a dose–response trend (Table 4). Both participants in the 4th MEHP quartile were statistically significant associated with higher LTL compared to the referent lowest MEHP quartile, but the dose–response trend was found only in the young adults (Table 4). Moreover, there were positive statistically significant association of the 4th MBP quartile and MBzP quartile compared to their respective lowest referent quartile in the older age group (Table 4). Analyses stratified by sex or by tobacco smoking use, confirmed the association of MEHP with longer LTL in both sexes and in smokers and non-smokers. (Supplemental Table 1). Sensitivity analyses including GGT yielded results similar to those from the primary analyses (data not shown).
Discussion Several phthalates have been associated with cell proliferation and cancer. In this study we confirmed the hypothesis that phthalate may be positively associated with LTL. The association of MEHP and LTL remained significant after inclusion of several confounder variables. Sensitivity analyses with the inclusion of the information of GGT (data not shown) provided further evidence of the non-spuriousness of the association. The statistical significant positive association, independently form age, remained also when the analyses were stratified among age groups (20–39years old, 40–59years and 60years and older), by sex and smoking. Moreover, a positive association was found in the older age group between MBP and MBzP with longer LTL. Based on the result of the multivariate analyses of this cohort, participants in the third and fourth quartile of MEHP had, on average, longer telomere length (129 and 172base pair, respectively) compared to those in the lowest quartile, whereas telomere erosion by age leads on average to a loss of 16bps per year. Telomeric DNA is lost with each cell division until replicative senescence is reached (Harley et al., 1990; Chiu and Harley, 1997). In the presence of telomerase, telomere length is maintained allowing the cells to escape replicative senescence (Liu et al., 2004). Human TERT (hTERT) promoter is the most important regulatory element of telomerase expression and contains numerous binding sites for a variety of transcription factors, including both activators and repressors of hTERT (Liu et al., 2004). There are several potential underlying mechanisms that may explain the positive association of MEHP and telomere length through the induction of several transcription factors that act as regulators of hTERT.